Docker version

[jinheong]

Hello!
I'm new to SAIGE, and I would like to use docker version since it might be the easiest way to begin with.
Is there anyone having experience of using docker to executing SAIGE?

I use a Mac mini computer for GWAS analysis.
Initially, I docker-pulled saige:1.3.0 and did step1 and step2 on the example data(nfam_100_nindep_0_step1_includeMoreRareVariants_poly.bed files), where the association result file was not generated.

Then I used another version saige:1.0.0, which generated the association results for example data.
However, when I turned to my own data (.bed files), the step 1 (with nThread=4) just exited as the progress message reached the following:

[1] "Start reading genotype plink file here"
nbyte: 5293
nbyte: 5293
reserve: 21314349056

M: 4025373, N: 21171

I monitored the CPU usage of SAIGE, which could reach up to 700-800% (i.e., 7 or 8 threads) and then SAIGE exited.

I guessed it might be the version problem, so I again pulled an older version saige:0.36.3.1.
This time, the step1 (nThread=4) reached further as following:

[1] "Start reading genotype plink file here"
nbyte: 5293
nbyte: 5293
reserve: 21314349056

M: 4025373, N: 21171
size of genoVecofPointers: 11
setgeno mark1
setgeno mark2

and it just hanged there and eventually exited after around 30 min.
The CPU usage just stayed around 100% during the process.

So, I'm wondering whether the problem is in docker version?
Or is my computer's configuration not capable of performing SAIGE for such big file (20000+ samples and 4000000+ variants)?
Any comments or suggestion would be appreciated.